Sex in public and private settings among Latino MSM

Abstract

Latino men who have sex with men (MSM) constitute a group at high risk for HIV. Recent approaches to understanding sexual risk have emphasized the role that contextual factors can play in shaping behavior. This study examined sexual behavior of Latino MSM in private and public settings. First, a within-person comparison of behaviors performed in sexual encounters that occurred in public and private settings was conducted. Unprotected anal intercourse and other sexual behaviors involving anal stimulation were more common in private settings; group sex was more likely in public settings. Second, a between-person analysis compared sexual behaviors of MSM who went to three different types of public sex settings during the previous six months. The types were: 1) commercial sex venues (CSVs), which were defined as businesses with the function of providing a space where MSM can go to have sexual encounters, such as gay bathhouses; 2) commercial sex environments (CSEs), which were defined as businesses with another apparent function, but in which MSM sometimes have sexual encounters, such as gay bars or pornographic movie houses; and 3) public sex environments (PSEs), which were defined as free public areas where MSM can go to find partners for anonymous sex, such as parks or public restrooms. Results indicated that anal sexual behavior was most likely to occur in CSVs and least likely in PSEs, but the probability of unprotected anal intercourse was not found to differ among the three types of settings. Behavioral differences were discussed in terms of structural conditions related to privacy and safety, and psychological factors related to intimacy.

HIV continues to be diagnosed at an alarming rate among men who have sex with men (MSM) in the U.S., accounting for approximately two-thirds of newly identified cases among men (CDC, 2008a). Moreover, prevalence among Latino men is double that of non-Hispanic white men (CDC, 2008b). Such statistics point to the need to understand sexual behavior patterns and the circumstances in which sexual risk occurs among Latino MSM. The current study examined Latin American immigrant MSM’s sexual behavior and risk in public and private settings.

Types of public settings for casual sex include commercial venues and free public spaces (Parsons & Vicioso, 2005). Among commercial settings, Binson and colleagues (2001) distinguished between 1) venues with the apparent purpose of providing a place for MSM to have sexual encounters (e.g., gay bathhouses, saunas, sex clubs)—which we labeled commercial sex venues (CSVs), and 2) venues with another ostensible purpose (e.g., gay bars, adult bookstores, and pornographic movie houses)—which we labeled commercial sex environments (CSEs). Free public spaces where MSM meet for sexual encounters (e.g., “cruising areas,” parks, public restrooms) are labeled public sex environments (PSEs).

Research findings indicate that about half of men sampled from gay communities reported sex in a public place during the previous year (Binson et al., 2001; Frankis & Flowers, 2005). Characteristics of men who have sex in public settings have been examined, and the partner pool in such settings tends to include men who engage in more risky behavior than men who do not have sex in public places. Two studies of Latino MSM found greater likelihood of UAI in the last 30 days among those who went to PSEs (Díaz et al.,1999) or to either CSVs or PSEs (Díaz et al., 1996) than among men who did not go to such venues.

Controversy about the role of public sex venues in the spread of HIV has been marked since early in the epidemic (Disman, 2003). Although unprotected anal intercourse (UAI) is more likely to occur in private settings than in public settings (Woods et al., 2007), and the most commonly reported behaviors performed in public settings carry lower risk of transmission (Tewksbury, 2002; Van Beneden et al., 2002; Woods et al., 2007), high risk behavior does take place. A probability sample of 400 MSM leaving a bathhouse found that 11% reported UAI at that visit (Woods et al., 2007), and similar percentages found in samples from Portland, Oregon (Van Beneden, et al., 2002) and Los Angeles (Richwald et al., 1988). The latter study also noted greater likelihood of UAI among Latino men. A review of the limited quantitative research on PSEs produced an estimate of approximately 10% men reporting UAI in a PSE during the previous 12 months (Frankis & Flowers, 2005).

Individuals’ self-protective behaviors differ depending on the situation (Bajos & Marquet, 2000; Barta et al., 2007; Zea et al., 2009); therefore, effective prevention requires an understanding of behavioral practices that occur in different contexts. Research concerning sexual risk in public settings has focused largely on UAI, and scant attention has been given to other sexual behaviors. Such behaviors, however, may have important health consequences. Among MSM, syphilis, gonorrhea, and Chlamydia are found in the genitals, rectum, or pharynx—transmitted through oral-genital, oral-anal, or genital-anal sex (Marcus et al., 2006). Oral-anal sexual activity has also been associated with transmission of other conditions, including herpes simplex (Fengyi et al., 2006), intestinal parasites (O’Connell et al., 2008; Wiwanitkit, 2006), typhoid fever (Reller et al., 2003), and shigellosis (Aragón et al., 2007). Although these conditions vary in their frequency, severity, and responsiveness to treatment, any of these could be serious for HIV-positive individuals.

In the current study, we explored relationships between setting and a variety of sexual practices among Latino MSM. First, we tested for within-person differences in the occurrence of sexual behaviors in public and private settings. Second, we compared sexual practices of Latino MSM who went to the different public venues (PSEs, CSEs, and CSVs) during the previous six months.

Methods

Participants and procedures

This study was part of a project concerning contextual influences on sexual risk behavior among Latin American immigrant MSM in the New York City metropolitan area. Three growing groups of Latino immigrants are those from Brazil, Colombia, and the Dominican Republic. The total sample included 482 MSM from these three countries, but the analyses for this paper were based on a subset of men who reported sex in a public setting. A survey was administered using computer-assisted self-interview technology (A-CASI) with audio enhancement and touch-screen responding. Participants chose whether to take the survey in Portuguese, Spanish, or English.

Inclusion criteria included having been born in Brazil, Colombia, or the Dominican Republic, residing in the New York City metropolitan area, being at least 18 years of age, having had sex in the last six months, and ever having had sex with men. Participants received reimbursement of $50 and a $15 stipend to cover transportation costs. Mean time to complete the survey was approximately 60 minutes. Details about recruitment and procedures for the survey, as well as descriptive information about the sample, were presented in a previous paper (Zea, Reisen, Poppen, & Bianchi, 2009).

Measures

Native speakers of Spanish or Portuguese translated measures used in the survey from English to Spanish or Portuguese, and then other individuals back-translated the material to English, either as part of the current or previous projects. In addition, Spanish speakers from different countries reviewed the measures to ensure the universality of the Spanish used. The portions of the survey that were used in the current paper are described below.

Sexual encounters

We asked participants about their most recent and penultimate sexual encounters. Depending on their answers, they were asked about other encounters, so that all participants either spontaneously reported or were queried about encounters involving anal intercourse with and without a condom, a public and private setting, and the presence or absence of drugs. Participants could indicate that they had never experienced a particular type of encounter.

For all encounters, participants were also asked when and where the encounter occurred. Location was assessed with an initial question, “Where did you have sex?” Response options included: my home, this partner’s home, our home, someone else’s home, a hotel/motel, or another place. If participants chose another place, they were asked a follow-up question concerning location with response options: sauna, bar or club, park or other outdoor space, bathroom, public transportation, movies/theater, car or truck, or other. In this paper, we coded as a private location anyone’s home or a hotel or motel. Among public locations, we coded sauna as a CSV; bar, club, or movies/theater as a CSE; and park, other outdoor space, or bathroom as a PSE. We excluded from the analysis encounters for which we could not determine whether the sex actually occurred in public or private (i.e., those who reported sex in a “car” or “other” location), as well as two cases of sex on public transportation. Although this location is a free, public setting, it is not the same as the cruising areas that we have classified as PSEs.

A list of sexual behaviors was presented for all encounters, and participants were asked to indicate whether the behaviors were performed, with possible responses of yes (1), no (2), or I don’t recall (3). The items consisted of explicit behavioral descriptions, as can be seen in Table 1, which includes the English, Spanish, and Portuguese versions. Other questions concerning the encounters addressed the number of partners participating and the use of alcohol or drugs.

Table 1.

Exact wording of sexual practices in three languages

Sexual Practices	English	Portuguese	Spanish
Oral Sex:	You sucked your sexual partner’s penis.	Você chupou o pênis do seu parceiro sexual.	Le mamaste el pene a tu pareja sex.
	Your sexual partner sucked your penis.	Seu parceiro sexual chupou o seu pênis.	Tu pareja sexual te mamó el pene.
Manual stimulation:	You masturbated your sexual partner until he came.	Você masturbou o seu parceiro sexual até ele gozar.	Masturbaste a tu pareja sexual hasta que se vino.
	Your sexual partner masturbated you until you came.	Seu pareceiro sexual te masturbou até você gozar.	Tu pareja sexual te masturbó hasta que te viniste.
Anal stimulation-tongue:	You used your tongue or mouth in and around your sexual partner’s anus.	Você usou sua lingua ou boca dentro ou em volta do ânus/cú do seu parceiro sexual.	Usaste tu lengua o tu boca en el ano de tu pareja sexual y a su alrededor.
	Your sexual partner used his tongue or mouth in and around your anus.	Seu parceiro sexual usou a lingua ou boca dele dentro ou em volta do seu ânus/cú.	Tu pareja sexual usó su lengua o su boca en tu ano o alrededor de él.
Anal stimulation-finger/hand:	You inserted your finger into your sexual partner’s anus.	Você inseriu o seu dedo dentro do ânus do seu parceiro sexual.	Le metiste el dedo por el ano a tu pareja sexual.
	Your sexual partner inserted his finger into you anus.	Seu parceiro sexual inseriu o dedo dele dentro do seu ânus.	Tu pareja sexual te metió el dedo por el ano.
Anal intercourse:	You inserted your penis in your sexual partner's anus.	Você inseriu o seu pênis no ânus do seu parceiro sexual.	Le metiste el pene por el ano a tu pareja sexual.
	Your sexual partner inserted his penis into your anus.	Seu parceiro sexual inseriu o pênis dele dentro do seu ânus.	Tu pareja sexual te metió su pene por el ano.
Unprotected anal intercourse:	You wore a condom.	Você usou camisinha.	Te pusiste un condón.
	Your sexual partner wore a condom.	Seu parceiro sexual usou camisinha.	Tu pareja sexual se puso condón.

HIV status questions

HIV status was asked, with response options of positive, negative, or I don’t know. Seroconcordance with partner was determined by combining the participant’s report of his own serostatus with reports about the serostatus of the partner for any given sexual encounter. Using responses that indicated definite knowledge of partner’s status (i.e., I know he was positive; I know he was negative), we determined whether or not there was known seroconcordance, which was coded as yes (1) versus no or unknown as (0).

Results

There were no significant differences among Brazilians, Colombians, and Dominicans in the likelihood of having sex in a public setting or in using a specific type of public venue (i.e., PSE, CSE, CSV). Moreover, we did not find an effect of country of birth on UAI in a public venue. Therefore, we combined the three national origins for the analyses reported here.

The first research question concerned sexual behaviors performed in public and private settings. The sample (N=315) for these analyses included those participants who reported at least one sexual encounter in a private place and at least one in a PSE, CSE, or CSV. We selected the most recent public and private encounters for each of these participants. We anticipated that there would be more manual and oral sex practices reported for public venues, and more anal sex, including unprotected anal intercourse as well as manual and lingual stimulation of the anus, in private venues.

Using the McNemar test for within-person comparisons, we examined whether the probability of specific sexual behaviors differed in private and public settings (see Table 2). As can be seen, individuals were more likely to perform the following behaviors in their encounters that occurred in private than in their encounters that occurred in public settings: unprotected anal intercourse, any anal intercourse (regardless of condom use), and stimulation of the anus with a tongue or fingers. In addition, in their private encounters, the men were more likely to have established that there was seroconcordance with their partners.

Table 2.

Within-person comparison of sexual behaviors at most recent sexual encounter in public and private settings (n = 315)

Behavior	Percent reporting in public encounter	Percent Reporting in private encounter	McNemar test of difference (S statistic)
Oral sex	88.0	92.0	3.8
Manual stimulation	54.0	52.0	0.5
Anal stimulation-tongue	28.0	60.0	62.3***
Anal stimulation- finger/hand	45.0	63.0	24.5***
Anal intercourse	55.0	80.0	50.0***
Unprotected anal intercourse	14.0	23.0	7.8**
Group sex	33.0	9.0	58.7***
Alcohol Use	34.0	39.0	2.1
Drug Use	25.0	25.0	0.0
Known seroconcordance	8.0	31.0	65.8***

^*p < .05;

^**p < .01;

^***p < .0001

The second research question concerned differences in behaviors performed in different types of public venues: PSEs, CSEs, and CSVs. We hypothesized that in CSVs there would be more anal sex, more UAI, and less mutual masturbation than in PSEs or CSEs. We anticipated high amounts of oral sex in all three types of venues. We restricted this analysis to those participants who reported a sexual encounter in a PSE, CSE, or CSV in the previous six months (N=182), which constituted 38% of the full sample. If a participant reported more than one such encounter in that time period, we selected the most recent.

As anticipated, there was more manual stimulation of a partner’s penis in the PSEs and no difference in the proportion of respondents in the three types of venues who reported having oral sex, which was a common behavior across venues (see Table 3). Contrary to expectations, no differences were found in unprotected anal intercourse, but as hypothesized, other types of anal sex—including anal intercourse and lingual and manual stimulation of the anus—were most prevalent in CSVs, as were encounters involving multiple partners or drug use. Many of these behaviors were seen least frequently in PSEs and at intermediate levels in CSEs. There was more alcohol use reported in CSEs than in the other venues, presumably due to the fact that that category included bars. Follow-up analysis revealed that the greater drug use found in CSVs was due to differences in use of poppers in the three settings. The only other drug which occurred with sufficient frequency to merit further examination was marijuana, which showed similar levels of use in CSVs, CSEs, and PSEs.

Table 3.

Between-person comparison of sexual behaviors in three types of public settings (n = 182)

	Percent reporting in public encounter
Behavior	CSV (n= 61)	CSE (n=52)	PSE (n= 69)	χ2 (df = 2)
Oral sex	92	85	86	1.55
Manual stimulation	43	48	66	6.85*
Anal stimulation-tongue	41	26	16	10.22**
Anal stimulation- finger/hand	61	44	32	10.66**
Anal intercourse	70	40	46	12.03**
Unprotected anal intercourse	11	10	6	1.36
Group sex	57	37	22	17.53***
Alcohol Use	26	52	22	13.79***
Drug Use	39	29	17	7.76*
Known seroconcordance	5	6	10	1.54
Own serostatus-- positive or unknown	56	44	43	2.32

^*p < .05;

^**p < .01;

^***p < .001

Discussion

As in previous research with MSM (e.g., Binson et al., 2001; Parsons & Vicioso, 2005), sex in public settings was found to be a common practice. Nearly two-fifths of the total sample reported having at least one sexual encounter in a public setting during the previous six months. Thus, greater knowledge of sexual behavior in these settings is important, and the current study makes a contribution despite the following limitations. Because we did not directly sample men in the three types of settings, it is unclear to what extent the men in this study who reported sexual encounters in public venues were representative of all the men who go to such venues. Moreover, the targeted sampling methods used yielded a sample that is not necessarily representative of Brazilian, Colombian, and Dominican immigrant MSM in the New York metropolitan area. Deriving a representative sample from hidden populations is difficult, as many people are motivated to remain unidentified due to immigration status, HIV status, or sexual orientation. Another limitation is related to the sampling of sexual encounters: participants who did not report anal intercourse or sex in public settings in their two most recent encounters were then asked about these behaviors.

Consistent with previous studies (e.g., Woods et al., 2007), we found greater likelihood of UAI in encounters that occurred in homes or hotels, in comparison to those that occurred in public venues. Other behaviors involving some anal contact (anal intercourse with a condom, rimming, and fisting) were also more likely at home. Structural features seem to be partially responsible for the difference, as home settings afford greater privacy and safety than do many public places where sex occurs. Nudity, which allows for anal sexual acts, is more easily accomplished in private environments. In public settings where the threat of discovery and the potential negative consequences of arrest, harassment, or embarrassment are present, it is beneficial to be able to break away quickly—which is certainly more difficult if one is even partially undressed. Home or hotel locations also typically have beds or couches, which provide greater comfort during a variety of sexual practices.

These structural characteristics are relevant not only to differences between private and public settings, but also to discrepancies found in behaviors performed in the three types of public venues studied here. The threat of discovery is greatest in PSEs, which are the sites where manual sex was found most frequently and anal practices were found least frequently. Because bathhouses afford a greater measure of privacy, safety, and comfort than other public settings (Tewksbury, 2002), it is not surprising that we found a greater likelihood of anal intercourse, rimming, and fisting reported there.

It is interesting to note that we did not find a higher probability of UAI in bathhouses than in CSEs and PSEs. Although Binson and colleagues (2001) reported more UAI in CSVs than in PSEs in a within-sample comparison, most arguments for greater risk in CSVs come from comparisons of different samples, and therefore, the apparent difference may be due to other confounding factors. The absence of an effect of venue type on unprotected sex in the current study, despite the greater incidence of anal intercourse found in CSVs, may be due to the effectiveness of current efforts in bathhouses to present prevention messages and to make condoms easily available (Binson et al., 2005; Tewksbury, 2002).

Psychological features can also contribute to differences in behaviors in private and public settings. In public venues MSM typically find anonymous, novel partners, whereas sexual encounters in private are frequently between known or primary partners. The stronger emotional connection between known partners may lead to a sense of safety, as well as a desire for the physical intimacy associated with unprotected sex (Elford, Bolding, McGuire, Sherr, 2001; Poppen et al., 2004) and with other anal sexual practices. Furthermore, emotionally intimate partners are more likely to know each other’s serostatus (Poppen et al., 2004) and may manage risk through sero-sorting strategies rather than condom use (Kippax et al., 1993). Consistent with this, we found that partners in private settings were more likely both to have established seroconcordance and to have engaged in UAI than were those in public settings.

Sexual encounters involving more than two partners were more common in public than in private settings. Furthermore, group sex was reported more frequently in CSVs than in the other two types of public settings, with over half of encounters in bathhouses involving multiple partners. The structural conditions described above, as well as the presence of communal areas such as steam rooms (Tewksbury, 2002), help create hospitable conditions for group sex. In addition, the several-hour visits typical in CSVs (Woods et al., 2007), the pervasive goal among attendees of sexual novelty and adventure (Tewksbury, 2002), and the social atmosphere (Tewksbury, 2002) are all conducive to group encounters.

Although we found drug use most prevalent in CSVs, further analysis revealed that this effect stemmed from use of poppers. Previous research has noted greater use of methamphetamines, ecstasy, and hallucinogens among MSM who went to CSVs, and greater use of barbiturates among MSM who went to PSEs, in comparison to those who did not (Parsons & Halkitis, 2002). We asked a slightly different question, which addressed drug use specifically in conjunction with sexual activity. It is interesting to note that participants reported very little use of drugs other than poppers and marijuana. We suspect that this discrepancy is due to the several factors. The men in this sample of Latino immigrants did not have money to spend on more expensive recreational drugs, and poppers are inexpensive and readily available. Moreover, poppers enhance sexual pleasure and are conducive to rapid sexual encounters. The focus on club drugs as an important influence on sexual risk behavior was less relevant for the Latino MSM in this study than it might be for other segments of the MSM community.

This study provided information concerning behavioral patterns and sexual practices in three types of public settings, as well as in private settings. Such knowledge can enable greater specificity in the design of prevention programs. For example, the sizeable number of participants engaging in oral-anal sex in CSVs suggests that educational materials concerning associated health risks could also be beneficial, particularly for individuals living with HIV. Moreover, despite lower rates of UAI in public compared to private settings, the substantial proportion of men with positive or unknown serostatus found among those who have sex in public settings, in combination with the lack of awareness of the partner’s’ serostatus, indicate the need for condom use programs in such settings.